#ceph IRC Log

Index

IRC Log for 2012-06-28

Timestamps are in GMT/BST.

[0:00] <sagewk> elder: it's not really a read or write, so the distinction is somewhat arbitrary. but it's not a write because it doesn't modify the object (only communicates ephemeral state to the notifier), which makes it a read by default under the current framework
[0:00] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) has joined #ceph
[0:02] <elder> sagewk, you wrote that as I was listening to Sam explain the same thing. Thanks.
[0:02] <sagewk> damn these headphones!
[0:02] <elder> It's a read because it doesn't requrire any durable state to change.
[0:02] * s[X]_ (~sX]@eth589.qld.adsl.internode.on.net) has joined #ceph
[0:02] <elder> (I thought that was why you wore them.)
[0:13] * sjustlaptop1 (~sam@aon.hq.newdream.net) has joined #ceph
[0:13] * sjustlaptop (~sam@aon.hq.newdream.net) Quit (Quit: Leaving.)
[0:20] <sjustlaptop1> sagewk: wip_2664
[0:30] * The_Bishop__ (~bishop@f052100067.adsl.alicedsl.de) has joined #ceph
[0:32] * The_Bishop_ (~bishop@e179003129.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[0:42] * sjustlaptop1 (~sam@aon.hq.newdream.net) Quit (Ping timeout: 480 seconds)
[0:47] * elder (~elder@aon.hq.newdream.net) Quit (Quit: Leaving)
[1:13] <jluis> sagewk, gregaf, around?
[1:13] <gregaf> I am!
[1:13] <jluis> cool
[1:14] <jluis> does something like this work for you?
[1:14] <jluis> ubuntu@plana41:~/ceph/src$ ./ceph mon add e 127.0.0.1:6799
[1:14] <jluis> added mon.e at 127.0.0.1:6799/0
[1:14] <jluis> ubuntu@plana41:~/ceph/src$ ./ceph mon add f 127.0.0.1:0
[1:14] <jluis> port defaulted to 6789; mon f 127.0.0.1:6789/0 already exists
[1:14] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) has left #ceph
[1:14] <gregaf> that looks like what we're after, yep!
[1:15] <jluis> cool, pushing it somewhere
[1:15] <gregaf> err, I assume it does the same thing with "./ceph mon add f 127.0.0.1"
[1:15] <gregaf> ?
[1:16] <gregaf> (it should given how the constructors work, but given how we noticed this bug I'm a little paranoid)
[1:17] <jluis> yeah
[1:17] <jluis> is it not supposed to?
[1:17] <jluis> ubuntu@plana41:~/ceph/src$ ./ceph mon add g 127.0.0.1
[1:17] <jluis> port defaulted to 6789; mon g 127.0.0.1:6789/0 already exists
[1:17] <gregaf> that's exactly what it should do, yes
[1:17] <jluis> okay
[1:18] <gregaf> and presumably "./ceph mon add h 127.0.0.2" outputs "port defaulted to 6789; added mon.h at 127.0.0.2:6789/0"
[1:19] <jluis> I suppose so
[1:20] <jluis> the patch itself is independent from the ip part of the address
[1:20] <jluis> just pushed it to wip-2661
[1:20] <jluis> it's simple and straightforward
[1:21] <gregaf> no implications on the error checking?
[1:21] <jluis> went with the CEPH_MON_PORT as the default one, instead of CEPH_PORT_FIRST though
[1:21] <jluis> no
[1:21] <gregaf> cool; that was my only worry :)
[1:21] <jluis> we just reassign the port to the default one
[1:21] * lofejndif (~lsqavnbok@659AAA9SC.tor-irc.dnsbl.oftc.net) Quit (Quit: gone)
[1:21] <jluis> the check is done on the following couple of lines
[1:23] * The_Bishop_ (~bishop@f052102144.adsl.alicedsl.de) has joined #ceph
[1:27] * The_Bishop__ (~bishop@f052100067.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[1:35] * BManojlovic (~steki@212.200.241.3) Quit (Quit: Ja odoh a vi sta 'ocete...)
[1:38] * Tv_ (~tv@aon.hq.newdream.net) Quit (Ping timeout: 480 seconds)
[1:42] * aliguori (~anthony@cpe-70-123-145-39.austin.res.rr.com) Quit (Remote host closed the connection)
[1:53] <sagewk> i forgot about the debian changes: switching to libnss, moving mds into ceph-mds package, mount.ceph and cephfs in ceph-fs-common package. any final input?
[1:53] <sagewk> maybe ceph-fs instead of ceph-mds for the package name?
[1:55] <gregaf> where's ceph-fuse going to be?
[1:55] <sagewk> still in ceph-fuse
[1:56] <gregaf> oh, didn't remember it being broken out
[1:56] * Ryan_Lane1 (~Adium@dslb-094-223-083-144.pools.arcor-ip.net) has joined #ceph
[1:56] * Ryan_Lane (~Adium@dslb-094-223-083-144.pools.arcor-ip.net) Quit (Read error: Connection reset by peer)
[1:56] <sagewk> what do you think, ceph-fs or ceph-mds?
[1:57] <dmick> I like ceph-fs better
[1:57] <dmick> parallel with -common, and everyone confuses that at first
[1:57] <gregaf> I do too, but my concern is that when I hear a name like that I expect it to include all the FS stuff
[1:57] <gregaf> :/
[1:57] <jluis> just what I was about to say
[1:57] <gregaf> and we do want to do independent installs of the mds and fuse, etc
[1:57] <dmick> presumably it'll depend on ceph-fs-common?
[1:57] <sagewk> not actually, it's just the server side
[1:57] <sagewk> nothing but ceph-mds daemon
[1:57] <nhm> I agree with greg on that
[1:57] <gregaf> it shouldn't; the mds and mount.ceph/cephfs have nothing to do with each other
[1:58] <dmick> hm, ok
[1:58] <sagewk> i will recommends ceph-fuse and ceph-fs-common, tho
[1:58] <sagewk> +Recommends: ceph-fuse, libcephfs1, ceph-fs-common
[1:58] <dmick> in that case I think ceph-mds is a better name
[1:59] <nhm> man, this aging test is going to take forever if the throughput keeps slowing down. :(
[2:01] <sagewk> gregaf: new patch in wip-msgr-fix to look at
[2:01] <sagewk> yeah, ceph-mds it is... otherwise the client/server stuff gets confusing.
[2:01] <sagewk> hrm, although, ceph (server) is alread separate from ceph-common (client, admin tools, etc.).
[2:01] <sagewk> well, whatever.
[2:03] <gregaf> whoops???*blush*
[2:03] <gregaf> hurray for counting patterns and use of enums making those values easy to change!
[2:08] <joshd> if it's just the mds, shouldn't it only suggest the client-side packages, not recommend them?
[2:11] * bchrisman (~Adium@108.60.121.114) Quit (Quit: Leaving.)
[2:30] * The_Bishop__ (~bishop@e177091113.adsl.alicedsl.de) has joined #ceph
[2:35] * The_Bishop_ (~bishop@f052102144.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[2:38] * yoshi (~yoshi@p22043-ipngn1701marunouchi.tokyo.ocn.ne.jp) has joined #ceph
[2:39] * sagelap (~sage@8.sub-166-250-80.myvzw.com) has joined #ceph
[2:43] * The_Bishop_ (~bishop@f052103120.adsl.alicedsl.de) has joined #ceph
[2:46] * The_Bishop__ (~bishop@e177091113.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[2:50] * The_Bishop__ (~bishop@f052101008.adsl.alicedsl.de) has joined #ceph
[2:54] * The_Bishop_ (~bishop@f052103120.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[2:58] * The_Bishop__ (~bishop@f052101008.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[3:07] * The_Bishop__ (~bishop@f052096197.adsl.alicedsl.de) has joined #ceph
[3:08] * sagelap1 (~sage@43.sub-166-250-67.myvzw.com) has joined #ceph
[3:10] * sagelap (~sage@8.sub-166-250-80.myvzw.com) Quit (Ping timeout: 480 seconds)
[3:16] * sagelap1 (~sage@43.sub-166-250-67.myvzw.com) Quit (Ping timeout: 480 seconds)
[3:24] * sagelap (~sage@52.sub-166-250-65.myvzw.com) has joined #ceph
[3:40] * sagelap (~sage@52.sub-166-250-65.myvzw.com) Quit (Ping timeout: 480 seconds)
[3:41] * sagelap (~sage@90.sub-166-250-69.myvzw.com) has joined #ceph
[3:44] * renzhi (~renzhi@178.162.174.69) Quit (Ping timeout: 480 seconds)
[3:45] * renzhi (~renzhi@208.53.156.36) has joined #ceph
[3:46] * joshd (~joshd@aon.hq.newdream.net) Quit (Quit: Leaving.)
[3:48] * rturk (~rturk@ps94005.dreamhost.com) Quit (Quit: Coyote finally caught me)
[3:48] * rturk (~rturk@ps94005.dreamhost.com) has joined #ceph
[3:54] * The_Bishop__ (~bishop@f052096197.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[3:55] * sagelap (~sage@90.sub-166-250-69.myvzw.com) Quit (Ping timeout: 480 seconds)
[3:55] * The_Bishop__ (~bishop@e177089190.adsl.alicedsl.de) has joined #ceph
[3:55] * sagelap (~sage@145.sub-166-250-77.myvzw.com) has joined #ceph
[4:07] * alis (~wjcw@659AAA9ZH.tor-irc.dnsbl.oftc.net) has joined #ceph
[4:10] * The_Bishop__ (~bishop@e177089190.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[4:11] * sagelap (~sage@145.sub-166-250-77.myvzw.com) Quit (Ping timeout: 480 seconds)
[4:12] * chutzpah (~chutz@216.174.109.254) Quit (Quit: Leaving)
[4:13] * The_Bishop (~bishop@f052099007.adsl.alicedsl.de) has joined #ceph
[4:13] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[4:14] * Ryan_Lane (~Adium@dslb-094-223-090-139.pools.arcor-ip.net) has joined #ceph
[4:15] * The_Bishop_ (~bishop@e177089190.adsl.alicedsl.de) has joined #ceph
[4:18] * Ryan_Lane1 (~Adium@dslb-094-223-083-144.pools.arcor-ip.net) Quit (Ping timeout: 480 seconds)
[4:21] * The_Bishop (~bishop@f052099007.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[4:25] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[4:25] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[4:31] * sjustlaptop (~sam@24-205-39-1.dhcp.gldl.ca.charter.com) has joined #ceph
[4:35] * izdubar (~MT@c-71-198-138-155.hsd1.ca.comcast.net) has joined #ceph
[4:35] * _adjohn (~adjohn@50-0-133-101.dsl.static.sonic.net) has joined #ceph
[4:36] * adjohn is now known as Guest1495
[4:36] * Guest1495 (~adjohn@50-0-133-101.dsl.static.sonic.net) Quit (Read error: Connection reset by peer)
[4:36] * _adjohn is now known as adjohn
[4:37] * loicd (~loic@2a01:e35:2eba:db10:120b:a9ff:feb7:cce0) Quit (Quit: Leaving.)
[4:37] * loicd (~loic@magenta.dachary.org) has joined #ceph
[4:38] * The_Bishop__ (~bishop@f052102142.adsl.alicedsl.de) has joined #ceph
[4:40] * The_Bishop_ (~bishop@e177089190.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[4:41] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[4:41] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Read error: Connection reset by peer)
[4:42] * MarkDude (~MT@c-71-198-138-155.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[4:57] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[4:57] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[5:04] * joshd (~jdurgin@2602:306:c5db:310:1e6f:65ff:feaa:beb7) has joined #ceph
[5:12] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[5:12] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[5:17] * The_Bishop (~bishop@f052101046.adsl.alicedsl.de) has joined #ceph
[5:19] * The_Bishop__ (~bishop@f052102142.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[5:28] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[5:28] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Read error: Connection reset by peer)
[5:29] * The_Bishop_ (~bishop@e179003229.adsl.alicedsl.de) has joined #ceph
[5:32] * bchrisman (~Adium@c-76-103-130-94.hsd1.ca.comcast.net) has joined #ceph
[5:32] * The_Bishop (~bishop@f052101046.adsl.alicedsl.de) Quit (Ping timeout: 480 seconds)
[5:43] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[5:44] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[5:46] * dmick (~dmick@aon.hq.newdream.net) Quit (Quit: Leaving.)
[5:59] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[5:59] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[6:15] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[6:15] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[6:19] * mdxi_ (~mdxi@74-95-29-182-Atlanta.hfc.comcastbusiness.net) has joined #ceph
[6:19] * mdxi (~mdxi@74-95-29-182-Atlanta.hfc.comcastbusiness.net) Quit (Read error: Connection reset by peer)
[6:23] * Ryan_Lane1 (~Adium@dslb-094-223-090-139.pools.arcor-ip.net) has joined #ceph
[6:23] * Ryan_Lane (~Adium@dslb-094-223-090-139.pools.arcor-ip.net) Quit (Read error: Connection reset by peer)
[6:30] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[6:30] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[6:46] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[6:46] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[6:48] * alis (~wjcw@659AAA9ZH.tor-irc.dnsbl.oftc.net) Quit (Remote host closed the connection)
[6:52] * lotreck (~Adium@c-75-75-33-53.hsd1.va.comcast.net) has joined #ceph
[7:02] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[7:02] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[7:16] * morse (~morse@supercomputing.univpm.it) Quit (Remote host closed the connection)
[7:17] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[7:17] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[7:33] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[7:33] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[7:49] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[7:49] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[7:52] * sjustlaptop (~sam@24-205-39-1.dhcp.gldl.ca.charter.com) Quit (Ping timeout: 480 seconds)
[8:04] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[8:04] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[8:05] * loicd (~loic@magenta.dachary.org) Quit (Quit: Leaving.)
[8:06] * loicd (~loic@magenta.dachary.org) has joined #ceph
[8:11] * s[X]_ (~sX]@eth589.qld.adsl.internode.on.net) Quit (Remote host closed the connection)
[8:16] * loicd (~loic@magenta.dachary.org) Quit (Quit: Leaving.)
[8:19] * coredumb (~coredumb@ns.coredumb.net) has left #ceph
[8:20] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[8:20] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[8:20] * lotreck (~Adium@c-75-75-33-53.hsd1.va.comcast.net) Quit (Quit: Leaving.)
[8:24] * The_Bishop_ (~bishop@e179003229.adsl.alicedsl.de) Quit (Quit: Wer zum Teufel ist dieser Peer? Wenn ich den erwische dann werde ich ihm mal die Verbindung resetten!)
[8:35] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) has joined #ceph
[8:35] * sagelap1 (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Quit: Leaving.)
[8:52] * sagelap (~sage@cpe-76-94-40-34.socal.res.rr.com) Quit (Ping timeout: 480 seconds)
[9:11] * loicd (~loic@83.167.43.235) has joined #ceph
[9:31] * BManojlovic (~steki@91.195.39.5) has joined #ceph
[9:56] * verwilst (~verwilst@d5152FEFB.static.telenet.be) has joined #ceph
[9:58] * MarkS (~mark@irssi.mscholten.eu) Quit (Remote host closed the connection)
[9:58] * MarkS (~mark@irssi.mscholten.eu) has joined #ceph
[10:01] * izdubar (~MT@c-71-198-138-155.hsd1.ca.comcast.net) Quit (Quit: Leaving)
[10:20] * joshd (~jdurgin@2602:306:c5db:310:1e6f:65ff:feaa:beb7) Quit (Ping timeout: 480 seconds)
[10:35] * d1g1t4l (~user@4.Red-2-139-228.staticIP.rima-tde.net) has joined #ceph
[10:35] <d1g1t4l> hi
[10:36] <d1g1t4l> im having problems with POSIX locks and ceph
[10:37] <d1g1t4l> i take a lock on a file from a process and then i kill the process with a signal
[10:37] <d1g1t4l> i would expect the lock to be released but it isnt
[10:37] <d1g1t4l> the same code using ext4 works fine
[10:38] <d1g1t4l> are posix locks broken?
[10:53] * adjohn (~adjohn@50-0-133-101.dsl.static.sonic.net) has left #ceph
[10:59] * morse (~morse@supercomputing.univpm.it) has joined #ceph
[11:22] * yoshi (~yoshi@p22043-ipngn1701marunouchi.tokyo.ocn.ne.jp) Quit (Remote host closed the connection)
[11:27] * Ryan_Lane (~Adium@dslb-094-223-090-139.pools.arcor-ip.net) has joined #ceph
[11:27] * Ryan_Lane1 (~Adium@dslb-094-223-090-139.pools.arcor-ip.net) Quit (Read error: Connection reset by peer)
[11:42] * lofejndif (~lsqavnbok@9YYAAHN9S.tor-irc.dnsbl.oftc.net) has joined #ceph
[11:43] * renzhi (~renzhi@208.53.156.36) Quit (Quit: Leaving)
[12:19] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) has joined #ceph
[12:20] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) has left #ceph
[12:34] * mgalkiewicz (~mgalkiewi@toya.hederanetworks.net) Quit (Quit: Leaving)
[12:38] * The_Bishop (~bishop@2a01:198:2ee:0:706f:bd55:8efa:c494) has joined #ceph
[13:21] * nhorman (~nhorman@2001:470:8:a08:7aac:c0ff:fec2:933b) has joined #ceph
[13:44] * mtk (~mtk@ool-44c35bb4.dyn.optonline.net) Quit (Read error: Connection reset by peer)
[13:49] * mtk (~mtk@ool-44c35bb4.dyn.optonline.net) has joined #ceph
[14:06] * aliguori (~anthony@cpe-70-123-145-39.austin.res.rr.com) has joined #ceph
[14:09] * newguy (7aa6e797@ircip1.mibbit.com) has joined #ceph
[14:11] * newguy (7aa6e797@ircip1.mibbit.com) Quit ()
[14:32] * Ryan_Lane1 (~Adium@dslb-094-223-090-139.pools.arcor-ip.net) has joined #ceph
[14:32] * Ryan_Lane (~Adium@dslb-094-223-090-139.pools.arcor-ip.net) Quit (Read error: Connection reset by peer)
[14:36] * fghaas (~florian@91.119.87.169) has joined #ceph
[15:05] * BManojlovic (~steki@91.195.39.5) Quit (Quit: Ja odoh a vi sta 'ocete...)
[15:06] * s[X] (~sX]@ppp59-167-157-96.static.internode.on.net) has joined #ceph
[15:09] * BManojlovic (~steki@91.195.39.5) has joined #ceph
[15:16] * aliguori (~anthony@cpe-70-123-145-39.austin.res.rr.com) Quit (Quit: Ex-Chat)
[15:20] * aliguori (~anthony@cpe-70-123-145-39.austin.res.rr.com) has joined #ceph
[15:31] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) has joined #ceph
[15:33] * elder (~elder@c-71-195-31-37.hsd1.mn.comcast.net) has joined #ceph
[15:47] * plastics (~plastics@c-69-138-42-222.hsd1.tn.comcast.net) Quit (Ping timeout: 480 seconds)
[16:18] * mdxi_ is now known as mdxi
[16:21] * s[X] (~sX]@ppp59-167-157-96.static.internode.on.net) Quit (Remote host closed the connection)
[16:25] * plastics (~plastics@c-69-138-42-222.hsd1.tn.comcast.net) has joined #ceph
[16:30] * MarkDude (~MT@c-71-198-138-155.hsd1.ca.comcast.net) has joined #ceph
[16:32] * Ryan_Lane1 (~Adium@dslb-094-223-090-139.pools.arcor-ip.net) Quit (Quit: Leaving.)
[16:34] * Ryan_Lane (~Adium@dslb-094-223-090-139.pools.arcor-ip.net) has joined #ceph
[16:42] * lofejndif (~lsqavnbok@9YYAAHN9S.tor-irc.dnsbl.oftc.net) Quit (Quit: gone)
[17:02] * johnl (~johnl@2a02:1348:14c:1720:15aa:62af:5d99:6ece) Quit (Remote host closed the connection)
[17:16] * verwilst (~verwilst@d5152FEFB.static.telenet.be) Quit (Quit: Ex-Chat)
[17:18] * johnl (~johnl@johnl.ipq.co) has joined #ceph
[17:21] * brambles (brambles@79.133.200.49) Quit (Remote host closed the connection)
[17:22] * brambles (brambles@79.133.200.49) has joined #ceph
[17:26] * johnl (~johnl@johnl.ipq.co) Quit (Ping timeout: 480 seconds)
[17:27] * loicd (~loic@83.167.43.235) Quit (Quit: Leaving.)
[17:28] * BManojlovic (~steki@91.195.39.5) Quit (Quit: Ja odoh a vi sta 'ocete...)
[17:28] * joshd (~jdurgin@2602:306:c5db:310:1e6f:65ff:feaa:beb7) has joined #ceph
[17:28] * loicd (~loic@83.167.43.235) has joined #ceph
[17:29] * plastics- (~plastics@c-69-138-42-222.hsd1.tn.comcast.net) has joined #ceph
[17:34] * plastics (~plastics@c-69-138-42-222.hsd1.tn.comcast.net) Quit (Ping timeout: 480 seconds)
[17:36] * loicd (~loic@83.167.43.235) Quit (Ping timeout: 480 seconds)
[17:41] * johnl (~johnl@2a02:1348:14c:1720:c0ae:8290:f70c:9d25) has joined #ceph
[17:42] * aliguori (~anthony@cpe-70-123-145-39.austin.res.rr.com) Quit (Remote host closed the connection)
[17:44] <sage> gregaf, joshd: did the rbd lockbreaking bits get merged? 2556 still says needs review
[17:44] <sage> joshd: also, the design doc should go into doc/ somewhere so that 2279 can be closed
[17:47] * lofejndif (~lsqavnbok@9YYAAHOLT.tor-irc.dnsbl.oftc.net) has joined #ceph
[17:55] * cattelan_away is now known as cattelan
[17:57] * Tv_ (~tv@aon.hq.newdream.net) has joined #ceph
[17:59] <joshd> sage: gregaf was fixing some things with it still last I heard
[18:01] * brambles_ (brambles@79.133.200.49) has joined #ceph
[18:03] * loicd (~loic@62-193-54-78.stella-net.net) has joined #ceph
[18:03] <joshd> sage, gregaf: dmick brought up a good point about shared locking - we may want to add another field that has to match to successfully take the shared lock, so you can have one set of systems that should be using it share that 'usage', for e.g. a single vm with live migration, you would could add 'usage=qemu-instance-10', and if you specify a different usage, you can't lock it
[18:04] * loicd (~loic@62-193-54-78.stella-net.net) Quit ()
[18:06] <fghaas> hrm, I wonder if this is intentional: ceph health says "HEALTH_WARN 1 mons down, quorum 1,2" (expected, mon is firewalled off), but there's no trace of that in ceph -w on any of the other mons. anyone able to offer insight on that?
[18:06] * loicd (~loic@62-193-54-78.stella-net.net) has joined #ceph
[18:12] * aliguori (~anthony@32.97.110.59) has joined #ceph
[18:17] <elder> sage, please update http://tracker.newdream.net/issues/2261 if you agree it should be marked "cannot reproduce"
[18:17] * loicd (~loic@62-193-54-78.stella-net.net) Quit (Quit: Leaving.)
[18:17] * lotreck (~Adium@mobile-198-228-196-180.mycingular.net) has joined #ceph
[18:18] * joshd (~jdurgin@2602:306:c5db:310:1e6f:65ff:feaa:beb7) Quit (Quit: Leaving.)
[18:20] * loicd (~loic@62-193-54-78.stella-net.net) has joined #ceph
[18:22] * BManojlovic (~steki@212.200.241.3) has joined #ceph
[18:22] * loicd1 (~loic@62-193-54-78.stella-net.net) has joined #ceph
[18:22] * loicd (~loic@62-193-54-78.stella-net.net) Quit (Read error: Connection reset by peer)
[18:26] * loicd1 (~loic@62-193-54-78.stella-net.net) Quit ()
[18:34] <sagewk> fghaas: i think i fixed ceph -s/-w/status output to include the quorum info
[18:40] * nhorman (~nhorman@2001:470:8:a08:7aac:c0ff:fec2:933b) Quit (Ping timeout: 480 seconds)
[18:40] <sagewk> joshd,dmick: good idea
[18:41] * loicd (~loic@62-193-54-78.stella-net.net) has joined #ceph
[18:45] * nhorman (~nhorman@hmsreliant.think-freely.org) has joined #ceph
[18:49] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) Quit (Ping timeout: 480 seconds)
[18:49] * lotreck (~Adium@mobile-198-228-196-180.mycingular.net) Quit (Quit: Leaving.)
[18:55] <fghaas> sagewk: in 0.47.3 ceph -w gets stuck like this: "mon e3: 3 mons at {daisy=192.168.42.114:6789/0,eric=192.168.42.115:6789/0,frank=192.168.42.116:6789/0}" ... no word of quorum
[18:55] <fghaas> (sorry, stepped out for a bit -- dinner with the kids)
[18:56] <sagewk> with 0.48 that line will include the quorum membership. and you see log messages go by when the quorum changes, elections are called, etc.
[18:56] <gregaf> joshd: sagewk: yeah, I was going to squash it down some but I sort of dropped it because Josh wanted positional arguments and I didn't want to hack them in since I'm not primary maintainer ;)
[18:56] <gregaf> just haven't picked it up since he hacked in the positional arguments
[18:57] <sagewk> gregaf: k
[18:57] <fghaas> sagewk: sweet, thanks
[18:59] * loicd (~loic@62-193-54-78.stella-net.net) Quit (Ping timeout: 480 seconds)
[18:59] * plastics- (~plastics@c-69-138-42-222.hsd1.tn.comcast.net) Quit (Ping timeout: 480 seconds)
[19:00] <fghaas> also (unrelated), http://wiki.ceph.com/wiki/OSD_journal states that using the CLONE_RANGE ioctl is a "possible optimization". I guess that's outdated because afaict ceph-osd does do that now, when on btrfs. out of curiosity, does it only use that for journal writes, or for something else too?
[19:02] <sjust> it refers to the possibility of cloning the write data from the journal directly to the file at apply time
[19:02] <gregaf> fghaas: what makes you think it's doing that?
[19:02] <gregaf> I'm pretty sure it only uses CLONE_RANGE when cloning between objects or doing snapshots
[19:02] <sjust> we don't currently do that since it would require the journal to be on the same fs as the file, which is a bit of a performance killer
[19:03] <sjust> we also use clone_range to speed recovery of files which happen to have chunks in common with files that have already been recovered
[19:03] <sjust> *we do use
[19:03] <sjust> it's also exposed via librados
[19:04] <sjust> to allow light-weight cloning between objects with the same locator
[19:05] <fghaas> sjust/gregaf: gotcha. misinterpreted some stuff in FileStore::_detect_fs()
[19:05] <gregaf> d1g1t4l: you around? you put that on the mailing list, right?
[19:06] <fghaas> sjust: prime performance killer, at least in my naive tests, was using btrfs at all, as opposed to sticking to xfs :)
[19:06] <sjust> fghaas: indeed
[19:06] <sjust> hopefully that situtation will improve...
[19:06] <sjust> *situation
[19:07] <nhm> fghaas: interesting, what kind of tests?
[19:07] <fghaas> just ceph osd bench stuff, really simple
[19:07] * loicd (~loic@90.84.144.136) has joined #ceph
[19:08] <nhm> fghaas: I typically see btrfs start out better and then degrade faster over time.
[19:08] <fghaas> nhm, I'll be back later if you want to enquire further :) (bedtime story reading duties now)
[19:08] <nhm> fghaas: XFS was better for me with lots of SSDs in the same node though.
[19:08] <nhm> ah, by all means!
[19:09] * d1g1t4l (~user@4.Red-2-139-228.staticIP.rima-tde.net) Quit (Quit: Lost terminal)
[19:11] * chutzpah (~chutz@216.174.109.254) has joined #ceph
[19:11] <jluis> where are we having the stand-up today?
[19:11] * joshd (~joshd@aon.hq.newdream.net) has joined #ceph
[19:15] * loicd (~loic@90.84.144.136) Quit (Ping timeout: 480 seconds)
[19:48] <yehudasa> nhm: on what machine are you running rest-bench?
[19:53] <nhm> plana83 and plana85, in a continuous loop as part of aging tests.
[19:54] <nhm> it's on iteration 43 right now. rest-bench should start up again in about 15 mins.
[20:03] <nhm> yehuda, btw, s3-func tests took 28s at the beginning of the aging tests, now they are taking ~160s. :(
[20:04] <yehudasa> nhm: I don't remember the config, is that xfs or btrfs?
[20:04] <nhm> yehudasa: this set is with btrfs
[20:04] <yehudasa> nhm: but no autodefrag, right?
[20:04] <nhm> correct
[20:05] <nhm> osds are around 23% full.
[20:11] <yehudasa> nhm: are you running the master branch or stable?
[20:11] <yehudasa> (or next)?
[20:12] <nhm> next (0.47.2-521-g88c7629)
[20:12] <fghaas> ok so nhm, my perf tests were really naive (and i wasn't interested in absolute performance numbers at all, just relative performance of backend filesystems)
[20:12] <fghaas> 3 nodes, all kvm virtualized using virtual block devices running on spinners, cache=none. tested with osd bench against freshly intialized osds with ext3, ext4, xfs and btrfs. once with journal in a file, once with a disk image backed by tmpfs in the host, passed to the guest as a virtio device, to simulate a fairly fast journal device. so 2 x 4 = 8 tests in total.
[20:12] <fghaas> benchmark results, best to worst: xfs, ext4, btrfs, ext3. no change in the ranking with regards to the journal device, except with the faster journal device xfs and ext4 are farther out front
[20:13] <nhm> rest-bench is running again now
[20:14] <nhm> fghaas: interesting! that was with rados bench or just osd tell bench?
[20:15] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) has joined #ceph
[20:15] <fghaas> nhm, the latter. I was primarily interested in single-osd performance, rather than the whole cluster
[20:16] <fghaas> xfs with the slow journal even beat btrfs with the "fast" journal
[20:16] <fghaas> (this is all on a 3.2 kernel I should add)
[20:16] <nhm> huh, strange.
[20:19] <fghaas> the numbers were (all MB/s): 87 vs 126 (ext3), 100 vs 160 (btrfs), 145 vs 300 (ext4), 170 vs 350 (xfs) ... like I said, the absolute numbers mean nothing, but the relation is kinda interesting
[20:21] <fghaas> nhm: strange? how so?
[20:22] <sagewk> fghaas: one thing to note with osd bench is that it is appending to one big object, not writing separate objects.. so the metadata cost is much lower than rados bench
[20:22] <nhm> fghaas: hrm... I don't trust those big numbers if you were just writing out to a standard spinning disk. It looks like you were just measuring tmpfs.
[20:23] <fghaas> it's not a single disk, it's the virt host is running on a pretty fast raid with a big write cache
[20:23] <nhm> fghaas: ok, but it's a 60s test right?
[20:24] <fghaas> um, osd bench only writes 1G afaict...
[20:24] <nhm> ok, even worse!
[20:24] <nhm> fghaas: How big is your tmpfs journal?
[20:25] <fghaas> well as far as ceph is concerned it has no clue that it's on tmpfs, as it's just a virtio block device which does use direct I/O, but that being said it's 1G
[20:26] <nhm> fghaas: Ok. I'm curious if you'd see the same results with a longer running test.
[20:26] * yehudasa_ (~yehudasa@aon.hq.newdream.net) has joined #ceph
[20:27] <fghaas> so yeah, I don't at all contest that I'm probably just measuring how fast my host write cache performs. but the difference is still quite startling :)
[20:27] <fghaas> difference between backend fs's, that is
[20:27] <sagewk> fghaas: fwiw osd bench can writ eout as much/little as you want if you pass it the magic undocumented arguments
[20:28] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) Quit (Quit: Leaving.)
[20:28] <fghaas> sagewk: then tell me the magic arguments and I'll see to them no longer being undocumented :)
[20:28] <sagewk> osd bench <bytes> <write_size>
[20:29] <sagewk> defaults are 1gb 4mb
[20:29] * mikeryan (~mikeryan@aon.hq.newdream.net) Quit (Quit: leaving)
[20:29] <sagewk> fghaas: thanks :)
[20:31] <nhm> fghaas: if you end up running any new tests let me know what you find. I'm curious how it will compare to some of the tests I've been doing.
[20:38] <fghaas> sagewk: actually, the docs really don't say a word about "ceph osd tell" at all
[20:39] <sagewk> yeah...
[20:40] <fghaas> I'm having a bit of a wtf moment here as "ceph osd tell 0 bench", while it continues to return "ok", has zero bearing on anything seen in ceph -w anymore
[20:40] <sagewk> if you do 'ceph tell osd.0 bench' it will block and wait for a reply
[20:41] <fghaas> ah, more magic
[20:42] * aliguori_ (~anthony@32.97.110.59) has joined #ceph
[20:42] * aliguori (~anthony@32.97.110.59) Quit (Read error: Connection reset by peer)
[20:55] <fghaas> sagewk: it actually seems to be bench <write_size> <bytes>
[20:55] <fghaas> which is in reverse order of the log message, mildly confusing
[20:57] * brambles (brambles@79.133.200.49) Quit (Remote host closed the connection)
[20:57] * brambles_ (brambles@79.133.200.49) Quit (Remote host closed the connection)
[20:57] * brambles (brambles@79.133.200.49) has joined #ceph
[21:22] * fghaas (~florian@91.119.87.169) Quit (Ping timeout: 480 seconds)
[21:30] * adjohn (~adjohn@50-0-133-101.dsl.static.sonic.net) has joined #ceph
[21:40] * lofejndif (~lsqavnbok@9YYAAHOLT.tor-irc.dnsbl.oftc.net) Quit (Quit: gone)
[22:04] * JT (~john@astound-64-85-227-32.ca.astound.net) has joined #ceph
[22:05] * plastics (~plastics@c-69-138-42-222.hsd1.tn.comcast.net) has joined #ceph
[22:07] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) has joined #ceph
[22:09] * aliguori_ (~anthony@32.97.110.59) Quit (Ping timeout: 480 seconds)
[22:28] * nhorman (~nhorman@hmsreliant.think-freely.org) Quit (Quit: Leaving)
[23:06] * aliguori (~anthony@cpe-70-123-145-39.austin.res.rr.com) has joined #ceph
[23:18] <nhm> joshd: Are you around?
[23:22] * loicd (~loic@magenta.dachary.org) has joined #ceph
[23:31] * loicd (~loic@magenta.dachary.org) Quit (Quit: Leaving.)
[23:34] * dmick (~dmick@aon.hq.newdream.net) has joined #ceph
[23:34] * sjustlaptop (~sam@aon.hq.newdream.net) has joined #ceph
[23:35] * plastics (~plastics@c-69-138-42-222.hsd1.tn.comcast.net) Quit (Ping timeout: 480 seconds)
[23:52] * baobrien (~baobrien@76.77.236.0) has joined #ceph

These logs were automatically created by CephLogBot on irc.oftc.net using the Java IRC LogBot.